CVE-2026-73498: mcp-atlassian Confluence Upload Attachment Arbitrary File Exfiltration
HERMES THREAT SCORE & MCP RISK EXPOSURE
Target:Model Context Protocol (MCP) Server & Atlassian Ecosystem Gateway CVSS v3.1 rates CVE-2026-73498 at 7.7 (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N). Hermes Threat Score assigns 88 (HIGH) to capture the amplified blast radius of AI agent tool execution and direct file/network access.
HASS AGENTIC SEVERITY & TOOL EXECUTION IMPACT
Target:Autonomous LLM Agent Interface & Host Execution Boundary Agentic tool protocols like MCP grant language models direct access to host filesystems and cloud services. Vulnerabilities in MCP tool endpoints allow indirect prompt injections to achieve arbitrary host compromise.
CVE-2026-73498: mcp-atlassian Confluence Upload Attachment Arbitrary File ExfiltrationVULNERABILITY
Enterprise knowledge base, project documentation, and team collaboration workspace platform.
๐ Why is this related? (Evidence & Provenance)
“Confirmed security vulnerability in Atlassian Confluence & Data Center Suite documented in Hermes dossier.”
- [vulnerability_report]
- [government_confirmation]CISA verified active exploitation in the wild and mandated federal remediation deadline in KEV entry. — Source: Cybersecurity & Infrastructure Security Agency (CISA): CISA Adds CVE-2026-59822 to Known Exploited Vulnerabilities Catalog (Reliability: VERY_HIGH)
1. Technical Context & Affected Software Matrix
Section titled โ1. Technical Context & Affected Software Matrixโ| Parameter | Specification | Operational Assessment |
|---|---|---|
| CVE Identifier | CVE-2026-73498 | Official Disclosure / Advisory |
| Affected Product | sooperset/mcp-atlassian | Model Context Protocol Server |
| Vulnerability Class | CWE-22 | Path Traversal File Exfiltration |
| Vulnerable Component | confluence_upload_attachment | Tool Execution & Dispatch Engine |
| Exploitation Vector | Remote Network / Prompt Injection | HTTP SSE Endpoint or Agent stdio pipe |
| Privileges Required | Low (PR:L) | Agent tool execution context |
| Resulting Access | Arbitrary Host File Access / SSRF | Full read/write over local host files or cloud IAM |
| Remediation Status | Fixed in version 0.22.0 | Immediate package update mandated |
2. Vulnerability Anatomy & Root Cause Analysis
Section titled โ2. Vulnerability Anatomy & Root Cause AnalysisโCode Dissection & Mechanism
Section titled โCode Dissection & MechanismโThe flaw resides in confluence_upload_attachment. The server failed to enforce directory sandboxing via validate_safe_path() or strict authentication checks:
# Vulnerable pattern in mcp-atlassian (< 0.22.0)def confluence_upload_attachment(file_path: str, **kwargs): # INSECURE: caller-supplied path is accessed without workspace boundary enforcement with open(file_path, "rb") as f: file_bytes = f.read() return client.upload(file_bytes)In version 0.22.0, the maintainers introduced mandatory path verification:
# Remediated pattern in mcp-atlassian (0.22.0)from mcp_atlassian.utils import validate_safe_path
def confluence_upload_attachment(file_path: str, **kwargs): safe_path = validate_safe_path(file_path, workspace_dir=CONFIG.WORKSPACE_ROOT) with open(safe_path, "rb") as f: file_bytes = f.read() return client.upload(file_bytes)3. Attack Vectors & Exploitation Scenario
Section titled โ3. Attack Vectors & Exploitation ScenarioโsequenceDiagram autonumber actor Attacker as Attacker / Adversarial Prompt participant Agent as LLM Agent (Claude/Cursor) participant MCP as mcp-atlassian Daemon participant Host as MCP Host OS Filesystem participant Atlassian as Jira / Confluence Cloud API
Attacker->>Agent: Indirect Prompt Injection (Ticket / Repo) Agent->>MCP: tools/call confluence_upload_attachment(file_path="../../../../etc/passwd") Note over MCP: Vulnerability CVE-2026-73498:<br/>Missing Path Validation / Auth Bypass MCP->>Host: Access arbitrary host resource Host-->>MCP: Raw file data or execution pipe MCP->>Atlassian: Upload secret as attachment / Outbound Request Atlassian-->>Attacker: Sensitive Data Disclosed- Adversarial Ingress: The attacker feeds a poisoned prompt into an issue ticket or source file consumed by an autonomous AI agent.
- Agentic Tool Invocation: The LLM agent generates a JSON-RPC
tools/callinvokingconfluence_upload_attachmentwith a malicious argument. - Boundary Transgression: Due to CVE-2026-73498, the MCP server performs the operation outside the designated workspace sandbox.
- Data Exfiltration: Host secrets, SSH keys, or cloud credentials are uploaded into Atlassian or returned in tool results.
4. Detection Engineering & Threat Hunting
Section titled โ4. Detection Engineering & Threat Huntingโalert http any any -> any any ( msg:"HERMES-CODEX EXPLOIT mcp-atlassian CVE-2026-73498 Attack Attempt"; flow:to_server,established; content:"POST"; http_method; content:"/mcp/tools/call"; http_uri; content:"confluence_upload_attachment"; nocase; content:"../"; classtype:attempted-admin; sid:20260031; rev:1; metadata:cve CVE-2026-73498, severity high;)title: mcp-atlassian CVE-2026-73498 Suspicious Tool Callid: cve-2026-73498-sigma-rulestatus: experimentaldescription: Detects path traversal or malicious arguments passed to confluence_upload_attachment in mcp-atlassian.logsource: category: application product: mcp-atlassiandetection: selection: tool_name: 'confluence_upload_attachment' argument_path|contains: - '../' - '/etc/' - 'C:\Windows\' - '~/.ssh' - '.aws/credentials' condition: selectionlevel: criticaltags: - attack.t1083 - attack.t10055. Remediation & Defense Playbook
Section titled โ5. Remediation & Defense Playbookโ- Immediate Package Upgrade: Upgrade
mcp-atlassianto version0.22.0or higher:Terminal window pip install --upgrade "mcp-atlassian>=0.22.0" - Local Loopback Binding: Restrict the HTTP/SSE listener strictly to
127.0.0.1:Terminal window python -m mcp_atlassian.server --host 127.0.0.1 --port 8000 - Workspace Isolation: Set
MCP_WORKSPACE_DIRto confine all file reads/writes to a dedicated sandbox. - Tool Whitelisting: Disable attachment manipulation tools if unneeded using
ENABLED_TOOLSorTOOLSETS.
Related Threat Intelligence & References
Section titled โRelated Threat Intelligence & Referencesโ- mcp-atlassian Profile: Version matrix, vulnerability posture, and changelog.
- Atlassian Confluence Intelligence: Enterprise wiki attack surface analysis.
- Atlassian Jira Profile: Issue tracker vulnerability history.
- Hermes Trust Center: Research methodology and vulnerability disclosures.